Politics & Policy

Parsing the Plethora of Polls

Even Axelrod agrees that there are too many suspect polls.

The political obituaries for the Republican nominee were rolling in prior to the first big debate.

William Saletan of Slate: “He is toast.”

Lawrence O’ Donnell of MSNBC: “It’s over.”

Eric Alterman of The Nation: “Barring extraterrestrial intervention, the election’s over and [Democrats] won it.”

The problem for Democrats is that all those baseless predictions were made in September 2000, when pundits were predicting that George W. Bush was a sure loser. Gallup’s polls in mid-September had Al Gore beating Bush by four to six points. Another poll released in mid-September by Newsweek showed Gore leading Bush by 14 points among likely voters. The Newsweek poll proved to be spectacularly wrong, in part because it surveyed only 580 likely voters over two nights. One of them was a Friday, when pollsters say more Democrats are at home.

This year pundits cited pre-debate polls, such as Gallup’s, that showed President Obama leading Mitt Romney by six points. But even then the poll’s margin of error was plus or minus 3.5 points. The race was still very much up for grabs.

Now the polling shoe is mostly on the other foot, because Romney has picked up momentum in the wake of his strong debate performance. According to the Real Clear Politics average of public polls, he trails Obama by only one point nationwide, he’s tied with Obama in Colorado and Virginia, and he trails him by only three points in Ohio.

The seesaw nature of the endless number of polls using a variety of methodologies spurs the poll-obsessed nature of campaign coverage. Such an obsession on polls may be insalubrious, helping to freeze public perceptions of the candidates and contributing to the low voter turnout that everyone in the media claims to lament.

This year, two national organizations are doing daily tracking polls. Gallup, the granddaddy of pollsters, now has Obama leading by 49 percent to 46 percent. But, in 2008, the most accurate of all major polls in predicting the outcome of the presidential race was that of Scott Rasmussen. His Rasmussen Reports nightly tracking poll now shows Romney leading Obama by 49 percent to 47 percent.

Rasmussen uses a telephone prompting system rather than live interviewers to survey respondents, a controversial technique in the eyes of traditional pollsters. But an analysis by The Progressive Review found that Rasmussen polls were the most accurate overall during this year’s primaries, and his lower costs allow him to survey more people and reduce his margin of error to 2 percentage points, significantly below that of other pollsters.

In recent years, many election polls have overestimated the Democrats’ share of the vote. People like to tell pollsters they plan to fulfill their civic duty and vote. In reality, about one-third of the people who say that won’t show up on Election Day. The best polls figure out if people are engaged enough to be likely voters. Even with those efforts, polls often oversample groups such as blue-collar women, who end up not voting in high numbers. Turnout, which has fallen below half of all adults in some years, does matter.

There are other reasons that polls may tilt slightly toward Democrats. The late Warren Mitofsky, who developed exit polling for CBS News in the 1960s, believed that Democrats are more likely to respond to media polls than are Republicans, who may distrust the media because of their liberal bias. In addition, more than 90 percent of the people pollsters try to contact initially hide behind voicemail or otherwise refuse to answer. “This makes survey results more uncertain and should cause concern, caution, and, above all, humility in reporting the results,” Leo Bogart, a former president of the American Association for Public Opinion Research, once concluded.

Unfortunately, polls help dictate media coverage. Reporters overemphasize certain elements of a candidate’s performance — Obama’s shaky handling of the economy or Romney’s verbal flubs — if the polls turn against one or the other. Then, until the polls turn again, coverage conforms to the stereotype. ABC’s George Stephanopoulos, a key strategist in Bill Clinton’s 1992 campaign, once said that political scientists “talk about the bandwagon effect, that once a candidate gets in the zone, all of the coverage is good, almost no matter what happens, and when you’re out of the zone, even when you do things right, it goes against you.”

The polls are a prime mover in this bandwagon effect. But before they drive a candidate’s supporters to despair or allow the media to beat a stereotype to death, some basic consumer reporting would help. Reporters should tell us which polls have a good track record, which have been clunkers, how much pollsters “push” the undecideds to a candidate, and how much less reliable some individual state surveys are. To do anything less risks heightening the cynicism of voters who increasingly feel that they’re being manipulated by both parties and not told the whole story by the media.

I sat in the green room with David Axelrod, President Obama’s chief strategist, at CBS News in Washington yesterday before we both appeared on Face the Nation. One thing we heartily agreed on is that there are too many suspect polls and too little independent analysis of how polls come to their conclusions. “I’ve seen polls that show us up more than we really are, and polls that wrongly show us down six points,” he told me. “I’ll say what I said months ago: This is going to be a very close race.” When the Axelrod Institute of Politics at the University of Chicago opens next January, one of the things he hopes it studies is the impact of never-ending polling on our elections.

Sadly, the evidence from the 2012 election so far is that the public’s suspicion of polls, the media, and the political process in general is only being heightened by what we are seeing and hearing.

— John Fund is national-affairs columnist for NRO.


The Latest